Re: [PATCH 1/4] printk: Allocate kernel log buffer earlier

[Ingo Molnar <mingo@elte.hu>]

* Mike Travis <travis@sgi.com> wrote:

> Yinghai Lu wrote:

> > please check updated patch...
> >
> > with the memblock change, you don't need to change acpi SRAT handling etc any 
> > more.
> 
> I had to debug a weird ACPI -> Node mapping last week and the
> "improved" SRAT messages helped that considerably.  It was
> far easier to spot which Node didn't have the correct assignments.
> I'd submit that patch even without needing fewer (like 512 lines
> max instead of 4096 lines max) bytes in the log buffer.

I agree that better compressed output generally makes sense - you just need to solve 
the ugliness aspect of it. (or get Len's Acked-by to add that code to drivers/acpi/)

Nevertheless doing this via memblock is obviously more important, as it solves the 
early printk log overrun problem once and for all.

> Let's move on to far more important problems.

That's not the threshold for upstream inclusion though. (at least for patches that 
we process via the -tip tree)

If you add crap to a single hardware driver then only that hardware is affected, but 
if you change the way the printk buffer is allocated it is *very important*, because 
like every single kernel message is affected by it.

So we scale up our review threshold with the importance of the piece of code 
affected.

Thanks,

	Ingo